Full Stack & Systems Engineer (Product & Cloud Delivery) · Cape Town (UTC+2)
Writing code is the easy part now. Owning the outcome in production is the job.
Responsibility as a Service (RaaS). I take work from the first conversation through to a verified production release across the stack, covering React and mobile WebXR on the frontend, Node.js APIs and Valkey caches in the middle, and AWS infrastructure underneath.
62%: Of total engineering delivery and production releases at Inner Reality (63% of org delivery volume across 25 repositories and 61% of production releases)
"Nothing he has taken ownership of has needed chasing. Where many developers wait for direction, Luke identifies what needs to happen, takes responsibility for the outcome, and drives it to completion."
Three Pillars of Delivery
Product & Cloud Delivery
Taking commercial briefs from a blank repository to live production. That includes mobile WebXR and React interfaces, partner REST APIs built on tight deadlines, and atomic Valkey and Lua prize flows under live sportsbook traffic.
Agentic Delivery ("The Loom")
Running Claude Code, Codex, and Antigravity through a custom safety harness with local execution guards, task ledgers, and automated tests. This workflow sustained 63% of org volume across 25 repositories over six months while keeping every production release gated by human verification.
Systems Rescue & Hardening
Grounded in a strong QA and systems foundation. Tracking down native Android WebXR memory crashes (SIGSEGV), reconciling drifted OpenTofu cloud state alongside the CTO, and putting automated security checks in place before launch.
Engineering Dossier
Every system below is written around what was built and the outcome it delivered for clients, partners, and the team.
Partner API & Campaign Runtime (Node.js · TypeScript · REST APIs): Implemented and shipped an external partner REST API in ten days for a Global Motorsport Partner alongside live campaign execution and automated budget throttling, unblocking a flagship integration on a fixed commercial deadline while protecting campaign budgets.
Distributed Cache & Atomic Prize Engine (Valkey Cluster · Lua · DynamoDB): Implemented the CTO designed hash tag key pattern and atomic Lua EVAL scripts across cluster Valkey (CROSSSLOT safe) to gate promotional prize claims, eliminating race condition over awards under live Tier 1 Sportsbook traffic and keeping prize accounting exact.
3D & WebXR Campaign Studio (React · TypeScript · Three.js · WebXR): Built a guided four step spatial campaign studio with automated 3D asset validation, live previewing, and direct publishing, removing engineering as a launch bottleneck and onboarding self serve paying clients.
Cloud & Release Control Plane (OpenTofu · AWS CDK · GitHub Actions): Maintained and operated AWS infrastructure and automated deployment pipelines alongside the CTO across 61% of production estate releases, turning manual deployments into repeatable promotions that reach customers the same day without downtime.
Independent Projects: GG Anvil (https://goodgameanvil.com/) gaming community operations platform on AWS Serverless, and a local Whisper and Ollama meeting to work pipeline on Linux.
Let's talk.
Looking for Full Stack, Product, or Platform Engineering roles where commercial outcomes depend on someone owning the work from start to finish. Available for full time or contract work from Cape Town (UTC+2), working UK and European hours and available for US East Coast hours, with a four week notice period.